Organic semiconductor heterointerfaces containing bathocuproine
ABSTRACT The four organic–organic heterointerfaces formed by depositing the organic semiconductor bathocuproine on tris(8-hydroxy-quinoline)aluminum ( Alq 3 ), N,N′- diphenyl -N,N′- bis(1-naphthyl) -1,1′ biphenyl -4,4″ diamine (α-NPD), 4,4′-N,N′- dicarbazolyl-biphenyl (CBP), and copper phthalocyanine (CuPc) have been studied using ultraviolet photoelectron spectroscopy. The relative positions of the vacuum levels and highest occupied molecular orbital levels of the organics were measured at each interface. Within our experimental uncertainty of 100 meV, the vacuum levels were found to align at all four interfaces. © 1999 American Institute of Physics.
- SourceAvailable from: Seth Coe-Sullivan[show abstract] [hide abstract]
ABSTRACT: The luminescence of inorganic core-shell semiconductor nanocrystal quantum dots (QDs) can be tuned across much of the visible spectrum by changing the size of the QDs while preserving a spectral full width at half maximum (FWHM) as narrow as 30 nm and photoluminescence efficiency of 50% [Journal of Physical Chemistry B 101 (46) (1997) 9463] . Organic capping groups, surrounding the QD lumophores, facilitate processing in organic solvents and their incorporation into organic thin film light-emitting device (LED) structures [Nature 370 (6488) (1994) 354] . A recent study has shown that hybrid organic/inorganic QD-LEDs can indeed be fabricated with high brightness and small spectral FWHM, utilizing a phase segregation process which self-assembles CdSe(ZnS) core(shell) QDs onto an organic thin film surface [Nature 420 (6917) (2002) 800] . We now demonstrate that the phase segregation process can be generally applied to the fabrication of QD-LEDs containing a wide range of CdSe particle sizes and ZnS overcoating thicknesses. By varying the QD core diameter from 32 Å to 58 Å, we show that peak electroluminescence is tuned from 540 nm to 635 nm. Increase in the QD shell thickness to 2.5 monolayers (∼0.5 nm) improves the LED external quantum efficiency, consistent with a Förster energy transfer mechanism of generating QD excited states. In this work we also identify the challenges in designing devices with very thin (∼5 nm thick) emissive layers [Chemical Physics Letters 178 (5–6) (1991) 488] , emphasizing the increased need for precise exciton confinement. In both QD-LEDs and archetypical all-organic LEDs with thin emissive layers, we show that there is an increase in the exciton recombination region width as the drive current density is increased. Overall, our study demonstrates that integration of QDs into organic LEDs has the potential to enhance the performance of thin film light emitters, and promises to be a rich field of scientific endeavor.Organic Electronics. 01/2003;
- [show abstract] [hide abstract]
ABSTRACT: The integration of organic and inorganic materials at the nanometre scale into hybrid optoelectronic structures enables active devices that combine the diversity of organic materials with the high-performance electronic and optical properties of inorganic nanocrystals. The optimization of such hybrid devices ultimately depends upon the precise positioning of the functionally distinct materials. Previous studies have already emphasized that this is a challenge, owing to the lack of well-developed nanometre-scale fabrication techniques. Here we demonstrate a hybrid light-emitting diode (LED) that combines the ease of processability of organic materials with the narrow-band, efficient luminescence of colloidal quantum dots (QDs). To isolate the luminescence processes from charge conduction, we fabricate a quantum-dot LED (QD-LED) that contains only a single monolayer of QDs, sandwiched between two organic thin films. This is achieved by a method that uses material phase segregation between the QD aliphatic capping groups and the aromatic organic materials. In our devices, where QDs function exclusively as lumophores, we observe a 25-fold improvement in luminescence efficiency (1.6 cd A(-1) at 2,000 cd m(-2)) over the best previous QD-LED results. The reproducibility and precision of our phase-segregation approach suggests that this technique could be widely applicable to the fabrication of other hybrid organic/inorganic devices.Nature 01/2002; 420(6917):800-3. · 38.60 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The idea to form C60/CuPc dispersed nanoheterojunctions by photoexcitation of a mixture of C60 and CuPc nanoparticles has been realized. The electronic structure of the nanoparticles and dispersed nanoheterojunctions formed in the mixture has been characterized by UV‐Vis spectroscopy and the comparing with known experimental ultraviolet photoelectron spectra and theoretical models of electronic structure of these molecules. For the mixture of C60 and CuPc nanoparticles in toluene and their coating layer on the quartz substrate the band offsets of the edges of CuPc VB and lowest unoccupied molecular orbital (LUMO) of C60 band are ΔE=1.55 eV and 1.4 eV, respectively. These results show clearly the presence of C60/CuPc dispersed nanoheterojunctions in the solution and on the quartz surface.Fullerenes Nanotubes and Carbon Nanostructures - FULLER NANOTUB CARBON NANOSTR. 01/2005; 13(3):259-272.
Organic semiconductor heterointerfaces containing bathocuproine
I. G. Hilla)and A. Kahn
Department of Electrical Engineering, Princeton University, Princeton, New Jersey 08544
?Received 5 April 1999; accepted for publication 9 July 1999?
The four organic–organic heterointerfaces formed by depositing the organic semiconductor
naphthyl?-1,1? biphenyl-4,4? diamine ??-NPD?, 4,4?-N,N?-dicarbazolyl-biphenyl ?CBP?, and
copper phthalocyanine ?CuPc? have been studied using ultraviolet photoelectron spectroscopy. The
relative positions of the vacuum levels and highest occupied molecular orbital levels of the organics
were measured at each interface. Within our experimental uncertainty of 100 meV, the vacuum
levels were found to align at all four interfaces. © 1999 American Institute of Physics.
Prior to the use of ultraviolet photoelectron spectroscopy
?UPS? to investigate molecular-level alignment at organic–
organic heterointerfaces, it was widely assumed that the
vacuum levels of two organic semiconductors would align at
the interface. When two materials are brought into contact,
the Fermi levels must align if the system is in thermody-
namic equilibrium. In general, vacuum-level alignment,
therefore, requires that the Fermi level be free to move in at
least one of the two materials, because of the different posi-
tions of the Fermi levels relative to the vacuum in the two
isolated materials. In the case of organic molecular semicon-
ductors, vacuum-level alignment was justified by citing the
closed-shell nature of the individual molecules, the resulting
weak intermolecular bonding, and the lack of interface states
within the semiconducting gaps. The Fermi level would be
free to move, and the vacuum levels should align. In such a
case, the offset between the two organic highest occupied
molecular orbitals ?HOMOs? would simply be equal to the
difference between the ionization energies of the two organ-
Recent UPS studies have confirmed that the vacuum lev-
els do align at the majority of organic heterointerfaces.1–5
Several exceptions have been found, including the interfaces
between perylenetetracarboxylic dianhydride ?PTCDA? and
(Alq3),6Alq3 and N,N?-diphenyl-N,N?-bis?1-naphthyl?-
1,1? biphenyl-4,4? diazine ??-NPD?,2and tetracyanoquin-
odimethane ?TCNQ? and tetrathianaphthacene ?TTN?.3The
formation of a charge-transfer complex, with an electron be-
ing donated by the TTN to the acceptor TCNQ, was used to
explain the TCNQ–TTN interface dipole.
Such donor–acceptor complexes may be expected at in-
terfaces between materials with large differences in electron
affinities. However, this cannot explain all the cases which
have been investigated. As noted above, the PTCDA/Alq3
interface exhibits a measurable dipole, while PTCDA/?-
NPD does not, despite the smaller electron affinity of ?-NPD
compared to Alq3.2,7
In this article, we report the results of an investigation of
four organic–organic heterointerfaces. The interfaces were
formed by depositing bathocuproine ?BCP? on Alq3, ?-NPD,
4,4?-N,N?-dicarbazolyl-biphenyl ?CBP?, and copper phtha-
locyanine ?CuPc?. The molecular structures of these materi-
als are presented in Fig. 1. In addition to the fundamental
importance of such a systematic study of interfaces between
one organic and a range of others, the first three of the above
interfaces have recently been used in organic light-emitting
devices ?OLEDs?,8,9and our results, therefore, have obvious
All experiments were performed in ultrahigh vacuum
?UHV?. Organics were deposited in a preparation chamber
?base pressure 4?10?10Torr?, which is connected to the
main analysis chamber ?base pressure 4?10?11Torr?. The
UPS system consists of a double-pass cylindrical mirror ana-
lyzer and a He discharge lamp. He I ?21.2 eV? and He II
?40.8 eV? photon energies were used. The overall resolution
of the system was estimated to be 150 meV from the width
of the Fermi-level step on freshly deposited Au.
Samples were prepared by depositing ?100 Å of a base
organic (Alq3, ?-NPD, CBP, or CuPc? on a flat substrate
?Si?100?:300 Å Cr:1000 Å Au?. All film thicknesses were
determined by timed depositions calibrated using a quartz-
crystal microbalance. No corrections for sticking coefficients
different from unity were used. The base organic was studied
using UPS to determine the quality of the film, to measure its
ionization energy, and to determine the binding energy of the
HOMO with respect to the Fermi level. The position of the
vacuum level was determined from the low-energy onset of
photoemission with the sample at ?3 V bias with respect to
the analyzer. Incremental thicknesses of BCP were deposited
on the base organic, doubling the total overlayer thickness at
each step. UPS spectra were collected at each thickness until
the data showed no contribution from the base layer, indicat-
a?Corresponding author. Electronic mail: firstname.lastname@example.org
JOURNAL OF APPLIED PHYSICS VOLUME 86, NUMBER 8 15 OCTOBER 1999
45150021-8979/99/86(8)/4515/5/$15.00© 1999 American Institute of Physics
ing a saturated overlayer with thickness greater than the
probing depth of the UPS technique. The data were then
compiled, examined for evidence of overlayer-induced
molecular-level displacement, hereafter called ‘‘bandbend-
ing,’’ and the HOMO-level offset and interface dipole mag-
nitudes were extracted.
The position of the top of the HOMO was estimated
from the zero crossing of the linear extrapolation of the low-
binding-energy side of the HOMO spectral peak. In the case
of BCP, there appear to be three overlapping peaks which
cannot be resolved and form a single, broad spectral feature.
This is illustrated in Fig. 2. The distance between the maxi-
mum of this feature and the top of the HOMO, as defined
above, is 1.75 eV. This fact was used to determine the posi-
tion of the top of the BCP HOMO at intermediate overlayer
thicknesses on heterointerfaces, where the overlap of spectra
from two organics obscured the low-binding-energy edge.
The position of the lowest unoccupied molecular orbital
?LUMO? has been estimated in each case by adding the en-
ergy of the onset of optical absorption to the HOMO level.
The position of the transport LUMO will differ from this
position for two reasons: ?1? the transport LUMO energy will
be increased by the electron–electron Coulomb interaction
energy, since the electron transport process involves adding
an extra electron to a neutral molecule; and ?2? the transport
LUMO energy will be decreased by molecular relaxation,
assuming the residence time of an electron on a molecule is
longer than the relaxation time. Although we do not know
the exact magnitude of these effects, they are expected to be
similar in both organics involved, such that the LUMO–
LUMO offset remains ? to that estimated using the optical
III. RESULTS AND DISCUSSION
The UPS data corresponding to the interface formed by
depositing BCP on Alq3are presented in Fig. 3. The bottom
spectrum is characteristic of clean Alq3, as has been reported
elsewhere.6The ionization energy of the Alq3was measured
to be 5.8?0.1eV, in excellent agreement with earlier work.6
The initial deposition of 4 Å BCP on the clean Alq3surface
results in a rigid shift of 50 meV towards higher binding
energies, indicating a slight overlayer-induced bandbending
in the Alq3layer. After deposition of 16 Å of BCP, the UPS
spectra resemble those of BCP on Au, and the Alq3features
FIG. 1. Molecular structures of the organics used in this study.
FIG. 2. He I ?21.22 eV? UPS spectrum of 64 Å BCP on Au. The lowest-
binding-energy feature appears to consist of three peaks ?fitted peaks shown
below?. The distance from the feature maximum to the top of the HOMO is
1.75 eV. Our definition of the top of the HOMO is illustrated in the inset.
FIG. 3. He I ?21.22 eV? UPS spectra of BCP on Alq3. Tick marks are
included as a guide to the eye. The inset shows the Alq3HOMO at low
coverages, illustrating the slight bandbending observed.
4516J. Appl. Phys., Vol. 86, No. 8, 15 October 1999I. G. Hill and A. Kahn
are completely supressed, indicating that the Alq3layer is
completely covered at this thickness. Further deposition of
BCP results in a slight ?100 meV? shift towards higher bind-
ing energies, which may be a result of bandbending in the
BCP layer, or may indicate charging within the wide-gap
material. The ionization energy of the BCP film was 6.4
?0.1eV. By removing the contributions of bandbending
from the HOMO–HOMO offset, the hole-transport barrier at
this interface was determined to be 0.65?0.1eV, with the
HOMO level of Alq3being above that of BCP, as illustrated
in Fig. 7. The LUMO levels, estimated using the optical
absorption gaps of 2.7 and 3.5 eV for Alq3?Ref. 10? and
BCP,10respectively, predict a small electron-transport barrier
of 0.15?0.1eV from Alq3to BCP. The onset of photoemis-
sion ?not shown? was used to determine the position of the
vacuum level at each thickness. The only movement ob-
served was attributed to the above-mentioned bandbending/
charging effects. The vacuum levels, therefore, align at this
interface, within our experimental uncertainty of ?0.1 eV.
This interface has recently been used in high-efficiency
electrophosphorescent OLEDs.8A thin layer of BCP was
inserted between two layers of Alq3—one used purely as an
electron-transport layer, and the other as a doped lumines-
cent layer. The purpose of the BCP was to confine holes
within the luminescent layer, which is reasonable consider-
ing the 0.65 eV hole barrier measured here. Additionally, for
this scheme to be efficient, electrons must be transported
from Alq3, across the BCP layer, and into the Alq3lumines-
cent layer. This requires near-alignment between the Alq3
and BCP LUMO levels at the interfaces. The LUMO–
LUMO offset estimated above is very small, and therefore,
consistent with the measured device performance.
Depositing BCP on the clean surface of ?-NPD does not
result in any appreciable bandbending within the ?-NPD
layer, as shown in Fig. 4. The spectral features of the clean
?-NPD surface, and the measured ionization energy of 5.4
?0.1eV are in excellent agreement with previously reported
values.1Increasing BCP thickness results in the attenuation
of the ?-NPD HOMO spectral intensity. It is difficult to
identify the BCP HOMO at intermediate thicknesses, be-
cause of the similarity in shape between the BCP and ?-NPD
features at the BCP HOMO binding energy. However, at 16
Å BCP, where the ?-NPD HOMO is completely attenuated,
we can unambiguously identify the BCP HOMO. Further
deposition of BCP results in a 200 meV shift of the spectral
features towards higher-binding energy, much as was ob-
served for BCP on Alq3. The ionization energy of the BCP
film was 6.5?0.1eV. Once again, the vacuum levels align at
the interface, within 0.1 eV. After subtracting the contribu-
tion from BCP bandbending, the HOMO–HOMO offset at
the interface was found to be 1.0?0.1eV, with the ?-NPD
HOMO above that of BCP ?Fig. 7?. Using the optical absorp-
tion band gaps ?3.1 eV for ?-NPD ?Ref. 11??, the electron-
transport barrier ?LUMO–LUMO offset? was estimated to be
0.6?0.1eV, with the LUMO of ?-NPD above that of BCP.
The UPS spectrum of clean CBP is shown at the bottom
of Fig. 5. The ionization energy of the clean CBP film was
6.0?0.1eV, as previously reported.12A slight ??100 meV?
shift of the CBP features was observed upon deposition of 4
Å of BCP. The CBP spectral features are not completely
supressed until a BCP thickness of 64 Å is reached, indicat-
ing that the BCP has a lower sticking coefficient on CBP,
that is, does not undergo layer-by-layer growth, or that the
initial CBP surface was rough, resulting in CBP protruding
through the BCP film until the higher film thickness was
reached. Little BCP bandbending ??100 meV? was ob-
served, and the ionization energy of the BCP film was mea-
sured to be 6.4?0.1eV, in agreement with the previous two
interfaces. The HOMO–HOMO offset at the interface is
0.44?0.10eV, with the CBP HOMO above that of BCP?Fig.
7?. The LUMO–LUMO offset, estimated using the optical
FIG. 4. He I ?21.22 eV? UPS spectra of BCP on ?-NPD. Tick marks are
included as a guide to the eye. The inset shows the ?-NPD HOMO position
at low coverages.
FIG. 5. He I ?21.22 eV? UPS spectra of BCP on CBP. Tick marks are
included as a guide to the eye. The inset shows the CBP HOMO position at
4517 J. Appl. Phys., Vol. 86, No. 8, 15 October 1999I. G. Hill and A. Kahn
band gaps ?3.1 eV for CBP ?Ref. 10??, is 0.04 eV, which is
less than our estimated uncertainty of 100 meV. The LUMO
levels, therefore, align with our experimental uncertainty.
Once again, the vacuum levels were found to align at this
The BCP/CBP interface has also been used in high-
efficiency OLEDs.8In this case, the BCP layer was inserted
between Alq3and the electron-transport layer, and a doped
CBP luminescent layer. Again, the purpose of the BCP was
to confine holes within the luminescent layer, consistent with
our measured hole barrier at this interface. Once again, for
this scheme to be efficient, the BCP and CBP LUMO levels
must align to allow electron transport into the CBP layer.
The LUMO–LUMO offset estimated above is negligible,
and therefore, consistent with the measured device perfor-
The results of the deposition of BCP on CuPc are pre-
sented in Fig. 6. The spectrum of clean CuPc, and its mea-
sured ionization energy of 5.2?0.1eV, agree well with pre-
vious studies.1The deposition of 4 Å BCP results in an
?200 meV overlayer-induced bandbending towards higher-
binding energy within the CuPc film. As in the case of BCP
on CBP, the CuPc features were not completely supressed
until a thickness 64 Å was reached. This behavior, indicating
a low sticking coefficient, or possibly the growth of islands
which coallesce at higher coverage, has been observed at
several interfaces containing CuPc.1Substantial BCP band-
bending of ?300 meV towards higher-binding energy was
observed with increasing thickness. The larger degree of
bandbending observed here may support the island growth
mechanism of BCP on CuPc, as the 64 Å spectrum may,
therefore, be sampling the top of islands which are substan-
tially thicker than the nominal coverage, and therefore, ex-
hibit a greater degree of bandbending ?or charging?. The ion-
ization energy of the saturated BCP film was measured to be
6.5?0.1eV. After correcting for these effects, the HOMO–
HOMO offset at the interface was determined to be 1.28
?0.1eV, with the CuPc HOMO well above that of BCP
?Fig. 7?. Using the optical gaps ?1.6 eV for CuPc ?Ref. 10??,
the LUMO–LUMO offset is 0.62?0.10eV, with the CuPc
LUMO below that of BCP. As was the case for the previous
three interfaces, the vacuum levels were found to align.
Interfaces between BCP and a range of different organ-
ics have been presented. These organics span a range of elec-
tron affinities from 2.3 eV ??-NPD? to 3.6 eV ?CuPc?, and
ionization energies from 5.2 eV ?CuPc? to 6.0 eV ?CBP?. It is
interesting that despite these wide ranges, the vacuum levels
were found to align at all interfaces. As far as the electron
affinities are concerned, this may be a result of the interme-
diate value of the BCP electron affinity ?3 eV?. The differ-
ence in electron affinities of the two interface constituents
does not exceed 0.7 eV at any of the four interfaces investi-
gated, in contrast to the reported electron affinity differences
between 0.8 and 1.5 eV at the Alq3/?-NPD, PTCDA/CuPc,
and PTCDA/Alq3interfaces. This justification appears less
reasonable, however, when ionization energies are consid-
ered, given that BCP has the largest ionization energy of the
group, 1.2 eV larger than that of CuPc. It is clear, in any
case, that the Fermi level at each of these organic surfaces is
able to move in an energy window sufficient to avoid the
formation of a compensating interface dipole.
UPS studies of four organic–organic heterointerfaces
have been reported. The four interfaces were formed by de-
positing BCP on Alq3, ?-NPD, CBP, and CuPc. The first
FIG. 6. He I ?21.22 eV? UPS spectra of BCP on CuPc. Tick marks are
included as a guide to the eye. The inset shows the CuPc HOMO at low
FIG. 7. Molecular-level diagrams summarizing the results of the four
organic–organic interfaces. The observed Fermi-level position at each inter-
face is indicated.
4518 J. Appl. Phys., Vol. 86, No. 8, 15 October 1999I. G. Hill and A. Kahn
three of these interfaces are used in OLEDs. The hole-
transport barrier ?HOMO–HOMO offset? was measured at
each interface. No dipoles were found at these interfaces,
implying that the vacuum levels align in all four cases.
Support of this work by the MRSEC program of the
National Science Foundation ?Award No. DMR-9809483?
and by the New Jersey Center for Optoelectronics ?Grant No.
97-2890-051-17? is gratefully acknowledged. One of the au-
thors ?I.H.? acknowledges support from NSERC of Canada.
The authors also thank the groups of S. R. Forrest and M. E.
Thompson for providing the organic materials.
1I. G. Hill and A. Kahn, J. Appl. Phys. 84, 5583 ?1998?.
2I. G. Hill and A. Kahn, Proc. SPIE 3476, 168 ?1998?.
3H. Ishii, K. Sugiyama, D. Yochimura, E. Ito, Y. Ouchi, and K. Seki, IEEE
J. Sel. Top. Quantum Electron. 4, 24 ?1998?.
4R. Schlaf, B. A. Parkinson, P. A. Lee, K. W. Nebesny, and N. R. Arm-
strong, Appl. Phys. Lett. 73, 1026 ?1998?.
5S. T. Lee, Y. M. Wang, X. Y. Hou, and C. W. Tang, Appl. Phys. Lett. 74,
6A. Rajagopal and A. Kahn, Adv. Mater. 10, 140 ?1998?.
7A. Rajagopal, C. I. Wu, and A. Kahn, J. Appl. Phys. 83, 2649 ?1998?.
8D. F. O’Brien, M. A. Baldo, M. E. Thompson, and S. R. Forrest, Appl.
Phys. Lett. 74, 442 ?1999?.
9G. Gu, G. Parthasarathy, P. E. Burrows, P. Tian, I. G. Hill, A. Kahn, and
S. R. Forrest, J. Appl. Phys. ?accepted?.
10S. R. Forrest ?private communication?.
11M. E. Thompson ?private communication?.
12I. G. Hill, A. Rajagopal, and A. Kahn, J. Appl. Phys. 84, 3236 ?1998?.
4519 J. Appl. Phys., Vol. 86, No. 8, 15 October 1999I. G. Hill and A. Kahn